The long term aim of our research is to develop and evaluate effective treatments for adolescents at high risk for suicide that will reduce suicidal and self-injurious behaviors as well as improve functioning and the quality of life of teens and their families. The primary aim of the research proposed here is to evaluate the efficacy of dialectical behavior therapy (DBT) for adolescents by comparing it to a combined individual and group supportive therapy control condition (I/GST) chosen specifically to maximize internal validity. Subjects will be 170 teens 13-17 years old with borderline personality traits referred to treatment due to high risk for suicide. Outcome targets include reduced frequency and severity of suicidal behaviors, especially suicide attempts, inpatient or ER treatment for suicidality, increased maintenance in and compliance with treatment, and enhanced functioning across multiple domains. DBT has an empirical track record with adults of reducing the incidence, frequency and medical risk of suicide attempts and non-suicidal self-injuries among individuals meeting criteria for borderline personality disorder (BPD). However, although DBT is widely used with suicidal adolescents, particularly those with difficulties characteristic of BPD such as poor emotion regulation and impulse control, no randomized trial of DBT with adolescents has been conducted. Thus, although non- randomized trials suggest effectiveness, without a randomized trial we simply do not know whether DBT for adolescents is efficacious or not. Given the severity of the problem and the lack of alternative treatments for high risk adolescents, addressing this question is important. The second aim of this research is to analyze mediators of reduced suicidal and self-injurious behaviors in adolescents. DBT is based on a theoretical model that suicidal behavior is a combined outcome of high stressful events and emotion dysregulation together with inadequate behavioral skills for coping with negative emotions and life events. Risk factors for adolescent suicide and intentional self-injury include most importantly family conflict but also conflict with peers, teachers, losses, legal or disciplinary problems (stressful events) combined with high negative emotions and inadequate abilities in emotion regulation. We will analyze the potential mediating effects on suicidal behaviors of decreases in family conflict, increases in parent DBT behavioral skills, reductions in emotion dysregulation and increases in DBT behavioral skills. Preliminary data on cost-effectiveness will also be collected.